Bladder cancer (BC) is a heterogeneous disease entity exhaling divergent phenotype, biological behavior and clinical outcome. About 70% of BC present as low-grade, superficial papillary tumors that are frequently recurrent but infrequently progress to high-grade muscle-invasive stages. The rest (~30%) of BC are high- grade and invasive at presentation, and they usually do have a prior history of low-grade, papillary tumors but are believed to arise de novo or derive from flat, high-grade carcinoma in situ lesions. Whether these two major phenotypic variants of BC are caused by distinct molecular alterations is of central importance in understanding and effectively managing BC. However, answer to this question remains elusive due to the lack of concerted research efforts. Over the past 15 years, we have been dissecting the molecular mechanisms of BC pathways by developing and analyzing genetically engineered mice via urothelium- specific gene activation or ablation or both. The resulting mouse models not only recapitulate many salient features of human BC, but yield new information regarding the cause-effect relationship between genes and pathways. The present proposal is designed to gain much deeper and broader insights into the combinatorial molecular events that together serve as the